Phase 1 Trial of Prophylactic Donor-Derived IL-2 Activated NK Cell Infusion after Allogeneic Hematopoietic Stem Cell Transplantation from a Matched Sibling Donor

Background: Donor lymphocyte infusion (DLI) is commonly used to enhance T-cell mediated graft-versus-tumor (GVT) effect after allogeneic stem cell transplantation (allo-HSCT). It was showed that prophylactic DLI can produce promising outcome but is also associated with a high rate of graft-versus-host disease (GVHD), limiting its use as maintenance treatment. By contrast with T cells, there are clinical and biological evidences that NK cells can provide strong GVT effect without inducing GVHD, making post allo-HSCT NK cell-based immunotherapy appealing. Thus, we designed a phase 1 clinical trial evaluating the safety of donor-derived ex vivo IL-2 activated NK cells (DLI-NK) as maintenance immunotherapy after allo-HSCT (DLI-NK: NCT01853358).

Methods: We included patients who were in morphologic remission of hematological cancers 2 months after allo-HSCT from a matched sibling donor after reduced intensity conditioning regimen, without history of acute GVHD. NK cells were purified from the donor (CD3 depletion / CD56 positive selection) and cultured ex vivo during 7 days with IL-2 before prior to infusion in patients between day+60 and day+90 after allo-HSCT. The study included a first phase of dose escalation (3+3 method) to determine the maximal tolerated dose (MTD) using 3 dose levels: 1.0 (dose 1), 5.0 (dose 2), 5.0-50 (dose 3) million cells/kg. A second phase allowed the treatment of subsequent patients to reach a total amount of 10 patients treated at the MTD. As ancillary studies, we performed: 1) preclinical evaluation of donor-derived IL-2 stimulated NK cell therapy in mice; 2) phenotypical and functional characterization of the NK cell therapy product before and after 7-day IL-2 stimulation, and 3) longitudinal immunomonitoring of NK cell phenotype in patients.

Results: We treated 16 patients without observing any dose-limiting toxicity (grade 3-4 adverse events related to DLI-NK within 30 days after infusion). During the follow up, 4 patients developed chronic GVHD (1 mild, 2 moderate and 1 severe forms). All of them were in complete remission of GVHD without immunosuppressive treatment at last contact. We observed one non-relapse death due to idiopathic pneumonia. From the four patients who relapsed (3 AML and 1 CLL), 2 died from their disease and 2 were in complete response after salvage therapy. At last follow up, 11 out of 16 treated patients were alive in CR without GVHD features and immunosuppressive treatment.

Ancillary studies showed that IL-2 stimulated allogeneic NK cells did not induce GVHD and displayed strong antitumor effect in mice against YAC-1 tumor cells, supporting the efficacy of this approach. In human, we observed that IL-2-activated NK cells expressed high levels of activating receptors (DNAM-1, NKG2A, NKp46, NKp30) and of the inhibitory receptor NKG2A, and exhibited increased degranulation and cytokine production when challenged by K562 tumor cells, as compared to freshly isolated NK cells. Immunomonitoring of patient NK cells after DLI-NK do not reveal significant phenotypic or functional dysfunction, with a pattern of NK cell recovery that is similar to what was previously observed after allo-HSCT.

Conclusion: We showed that prophylactic DLI-NK after matched sibling donor allo-HSCT was feasible and safe. We observed low incidence of GVHD as compared to previous reports of conventional prophylactic DLI, reinforcing the rationale of NK cell role in GVHD prevention. In addition, promising disease-free survival supports the design of future clinical trials evaluating the efficacy of DLI-NK as disease relapse prevention.